I took the liberty of including your attached schematic below. Perhaps you attached the wrong schematic? The following is not a constant lighting circuit; the brightness will vary with track voltage. Also the capacitor voltage (20V) is under-rated for O-gauge command voltage operation. I realize this thread is about low-voltage conventional so maybe your transformers can't reach command voltage levels. But higher voltage capacitors (e.g., 35V, 50V) are about the same price so irrespective of control circuitry used, if modifying an O-gauge AC-powered car for DC LED operation, I'd use a higher voltage capacitor.
Stan
I think that resistor is undersized. With the capacitor at maximum throttle there would be over 24 volts DC into the LED and resistor. You would need at least a 1000 ohm half watt resistor. 1200ohm would be better. If you install a Cl2-n3 chip instead of the resistor you would get CV lighting
Dale H
Note that the CL2 will be running pretty hot dropping 21 volts, the TO-92 package doesn't handle a lot of power well, temperature rise with that drop will be over 50C. I've done this on the bench, it gets REALLY hot.
Understood. I believe you're seeing a measurement artifact when using the AC-current function of some DMMs. I'm sure someone else can state it more simply but here's how I see it. If an AC voltage drives just a resistor, the AC current follows the voltage...double the voltage, double the current; if the voltage is a sinewave, so too will be the current. But with that bridge rectifier and capacitor, the cap gets charged up to some DC voltage and current only flow through the bridge when the input voltage exceeds the charged capacitor voltage. This occurs at the peaks of the AC signal so current from the transformer into the bridge is actually zero most of the time. In other words you can have a smooth AC sinewave voltage going into the bridge, but the AC current will be something different. It is this something different current signal that I suggest is causing the apparent measurement error. A discussion of how to accurately measure complex AC waveforms with the high-frequency electronic signals in modern O-gauge circuits can get somewhat nerdy.
Probably a lot nerdy!
This is why I sometimes test with DC for these measurements. I can measure the current accurately using DC.